Information Notice No. 82-26: RCIC and HPCI Turbine Exhaust Check Valve Failures

                                                       SSINS No.: 6835  
                                                       IN 82-26 

                               UNITED STATES 
                       NUCLEAR REGULATORY COMMISSION 
                    OFFICE OF INSPECTION AND ENFORCEMENT 
                           WASHINGTON, D.C. 20555 

                               July 22, 1982 

Information Notice No. 82-26:   RCIC AND HPCI TURBINE EXHAUST CHECK VALVE
                                   FAILURES 

Addressees: 

All boiling water nuclear power reactor facilities holding an operating 
license or construction permit. 

Purpose: 

This information notice is provided as an early notification of a 
potentially significant problem pertaining to reactor core isolation cooing 
(RCIC) and high pressure coolant injection (HPCI) turbine exhaust check 
valve failures. It is expected that recipients will review the information 
for applicability to their facilities. No specific action or response is 
required at this time. 

Description of Circumstances: 

A number of RCIC turbine exhaust check valve failures that have occurred 
during the past 20 months are outlined below. 

1.   On December 10, 1980, Carolina Power and Light Company reported (LER 
     80-101/03L) a RCIC system turbine trip at Brunswick Steam Electric 
     Plant, Unit 2, while conducting a RCIC system test. The turbine tripped
     on high turbine exhaust pressure due to the turbine exhaust swing check
     valve failing in the closed position. Inspection revealed the check 
     valve disc stem had broken off where it connects to the valve hinge 
     assembly. This allowed the disc to fall into the discharge part of the 
     valve and isolate flow. An examination of the check valve disc and 
     hinge assembly indicated the disc had been rotating inside of the hinge 
     bore area and caused excessive wear of both components. In addition, 
     indications that the valve disc had been striking the upper part of the 
     valve body while in the open position were noted. To return the check 
     valve to normal operability, the valve seat was lapped, the valve disc 
     replaced, and the valve was tested satisfactorily. 

     This LER also referred to a similar failure (LER 79-074/03L) at 
     Brunswick Steam Electric Plant, Unit 1. This time, disassembly of the 
     RCIC steam exhaust check valve showed that the stud and nut on the back
     of the disc had broken and the disc had separated from the hinge and 
     had lodged in the valve inlet. A new valve was ordered and installed 
     upon arrival. The valve failure prevented the RCIC turbine, which had 
     been used intermittently throughout the day for vessel level control, 
     from starting following a reactor scram. 


8204210398 
.

                                                           IN 82-26 
                                                           July 22, 1982  
                                                           Page 2 of 4 

2.   On May 29, 1981, Pennsylvania Power and Light Company reported (ERs 
     100450/100508) the failure of the RCIC turbine exhaust swing check 
     valve at Susquehanna Steam Electric Station, Unit 1, while conducting a 
     RCIC system test. The stud (integrally cast with the disc) which 
     attached the disc, to the valve hinge broke off. In a subsequent report 
     on February 5, 1982, they indicated that turbine exhaust steam flow 
     conditions experienced during testing caused the valve disc to cycle 
     violently open and close. Since the check valve was sized for full 
     flow, operational testing of the system at low flow caused the disc 
     function to be erratic. As a result, the end of the disc stud gradually 
     wore a hole in the valve bonnet (cover) which served as the stop. This 
     additional travel allowed the disc edge to impact against the valve 
     body due to a lack of clearance. The loads and stresses experienced by 
     the disc resulted in a disc stud fracture. The failure was a brittle 
     fracture. A second disc, taken from Unit 2, was put in service to 
     replace the fractured disc. This disc also failed at approximately the 
     same section as the first. Therefore, they concluded that with both the 
     valve and system as presently designed, a swing check valve disc will 
     fail for this service application. This was further evidenced by three 
     more replacement discs that eventually broke in a similar fashion in 
     spite of the provision of a specially designed "anvil" nut to replace 
     the original nut. They are planning on either replacing the existing 
     valve with a lift type check valve design having an inherent damping 
     action in the opening position or modifying the existing valve and/or 
     piping system so that the valve will function properly under both, low 
     and high flow conditions. (See Information Notice No. 82-20.) 

3.   On December 10, 1981, Georgia Power Company reported (LER 81-112/03L),a
     RCIC isolation at Edwin I. Hatch Nuclear Plant, Unit 2, while 
     conducting a RCIC rated flow test. An investigation revealed that the 
     turbine exhaust check valve had internal damage creating a block in the 
     line causing the rupture diaphragm to fail. The valve was repaired and 
     the diaphragm replaced. A design change has been approved to replace 
     the check valve with a better design. The new valve has been ordered 
     and will be installed as soon as possible. 

     A generic review, by the licensee, revealed that the HPCI system has 
     the same valve type in a similiar configuration and that a design 
     change has been approved to replace the valve. 

4.   On March 9, 1982, Long Island Lighting Company reported a deficiency 
     concerning two check valves located in the RCIC turbine exhaust line at
     Shoreham Nuclear Power Station. The deficiency was identified while t 
     testing the turbine and pump using auxiliary steam at low flow 
     conditions. Examination of the valves disclosed that the slamming and 
     cyclic action of the valve resulted in wear to the swing check 
     bushings, the anti-rotation pins, and the swing checks. The valve 
     bodies showed rubbing marks from the interaction with the swing check. 
     A systems review of the valve failures, by the licensee, indicated that 
     damage to these components could have an 
.

                                                           IN 82-26  
                                                           July 22, 1982  
                                                           Page 3 of 4 

     impact on the turbine exhaust back pressure thereby causing the turbine
     to trip. All damaged components of the RCIC exhaust check valves will 
     be replaced and the valves will be rebuilt to assure properly 
     conditioned and working valves are installed. 

Discussion: 

All of the above failures deal only with the RCIC turbine exhaust check 
valve. However, as noted by Georgia Power Company, the HPCI exhaust system 
has the same type valve in a similiar system configuration. Thus it is 
reasonable to expect similiar problems with the HPCI turbine exhaust check 
valve also. In fact, both services have been identified in the generic 
correspondence by General Electric pertaining to this topic. 

The first of the generic correspondence is Services Information Letter (SIL)
No. 30, "HPCI/RCIC Turbine Exhaust Line Vacuum Breakers," dated October 31, 
1973. In this SIL, General Electric identified the problem of possible 
damage to the exhaust line check valve and recommended the installation of 
vacuum breakers based on tests conducted at Browns Ferry and Peach Bottom. 

The second of the generic correspondence is Application Information Document
(AID) No. 56, "High Pressure Core Injection and Reactor Core Isolation 
Cooling Turbine Exhaust Check Valve Cycling," dated December 18, 1981. In 
this AID, General Electric identified the possible causes of failure as 
improper system operation, improper check valve sizing, inadequate check 
valve design, or inadequate exhaust line design. To minimize the possibility
of future problems, they recommend that: 

1.   Manual starts and monthly system surveillance testing should be 
     performed in accordance with the Operating and Maintenance Instructions
     (specifically, gradually increasing the turbine speed until the rated 
     pump discharge flow is achieved is not recommended). 

2.   The exhaust check valve, the exhaust line vacuum breaker, and the 
     exhaust line sparger should be designed in accordance with the 
     requirements/ recommendations given in the GE system design 
     specification. 

3.   System operation below the recommended turbine rated speed should be 
     minimized. 

4.   The exhaust check valve should be located as close as possible to the 
     containment. 

5.   The turbine exhaust check valve internals should be visually inspected 
     on a routine schedule such as at every refueling outage. 
.

                                                           IN 82-26  
                                                           July 22, 1982  
                                                           Page 4 of 4 

If you have any questions regarding this matter, please contact the Regional
Administrator of the appropriate NRC Regional Office, or this office. 



                              Edward L. Jordan, Director  
                              Division of Engineering and 
                                Quality Assurance 
                              Office of Inspection and Enforcement 

Technical Contact:  Richard J. Kiessel 
                    301-492-8119 

Attachment: 
List of Recently Issued IE Information Notices